US Navy tests Lockheed Martin AI sonar software for submarine detection

The US Navy tested Lockheed Martin’s SensorMAX AI-driven sonar processing system for submarine detection during exercise RIMPAC 2026 off the coast of Hawaii, according to a statement released by the defense contractor.

During the maneuvers, MH-60R Seahawk helicopters deployed sonobuoys that transmitted acoustic sensor data directly to the SensorMAX platform. The system uses machine learning algorithms to analyze acoustic feeds and issue real-time operational updates to anti-submarine warfare units.

Company representatives noted that when a new target is detected, system operators flag the acoustic signature, retrain the model using the fresh data, and transmit a small update package back to the aircrew. Each retraining iteration takes under five minutes to execute.

The platform allows crews to process up to eight concurrent sonobuoy data streams. Operating independently of specific hardware, SensorMAX can ingest and analyze acoustic inputs from air, surface, and ground sensors. Developers stated that this platform-agnostic classification capability enables naval forces to expand defensive perimeters around high-value assets and accelerate the anti-submarine kill chain.